Re: [PATCH bpf-next v2 22/35] bpf: eliminate rlimit-based memory accounting for bpf ringbuffer

2020-07-27 Thread Andrii Nakryiko
On Mon, Jul 27, 2020 at 12:21 PM Roman Gushchin  wrote:
>
> Do not use rlimit-based memory accounting for bpf ringbuffer.
> It has been replaced with the memcg-based memory accounting.
>
> bpf_ringbuf_alloc() can't return anything except ERR_PTR(-ENOMEM)
> and a valid pointer, so to simplify the code make it return NULL
> in the first case. This allows to drop a couple of lines in
> ringbuf_map_alloc() and also makes it look similar to other memory
> allocating function like kmalloc().
>
> Signed-off-by: Roman Gushchin 
> ---

LGTM.

Acked-by: Andrii Nakryiko 

>  kernel/bpf/ringbuf.c | 24 
>  1 file changed, 4 insertions(+), 20 deletions(-)
>

[...]


Re: [PATCH bpf-next v2 22/35] bpf: eliminate rlimit-based memory accounting for bpf ringbuffer

2020-07-27 Thread Song Liu
On Mon, Jul 27, 2020 at 12:22 PM Roman Gushchin  wrote:
>
> Do not use rlimit-based memory accounting for bpf ringbuffer.
> It has been replaced with the memcg-based memory accounting.
>
> bpf_ringbuf_alloc() can't return anything except ERR_PTR(-ENOMEM)
> and a valid pointer, so to simplify the code make it return NULL
> in the first case. This allows to drop a couple of lines in
> ringbuf_map_alloc() and also makes it look similar to other memory
> allocating function like kmalloc().
>
> Signed-off-by: Roman Gushchin 

Acked-by: Song Liu 

> ---
>  kernel/bpf/ringbuf.c | 24 
>  1 file changed, 4 insertions(+), 20 deletions(-)
>
> diff --git a/kernel/bpf/ringbuf.c b/kernel/bpf/ringbuf.c
> index e8e2c39cbdc9..e687b798d097 100644
> --- a/kernel/bpf/ringbuf.c
> +++ b/kernel/bpf/ringbuf.c
> @@ -48,7 +48,6 @@ struct bpf_ringbuf {
>
>  struct bpf_ringbuf_map {
> struct bpf_map map;
> -   struct bpf_map_memory memory;
> struct bpf_ringbuf *rb;
>  };
>
> @@ -135,7 +134,7 @@ static struct bpf_ringbuf *bpf_ringbuf_alloc(size_t 
> data_sz, int numa_node)
>
> rb = bpf_ringbuf_area_alloc(data_sz, numa_node);
> if (!rb)
> -   return ERR_PTR(-ENOMEM);
> +   return NULL;
>
> spin_lock_init(>spinlock);
> init_waitqueue_head(>waitq);
> @@ -151,8 +150,6 @@ static struct bpf_ringbuf *bpf_ringbuf_alloc(size_t 
> data_sz, int numa_node)
>  static struct bpf_map *ringbuf_map_alloc(union bpf_attr *attr)
>  {
> struct bpf_ringbuf_map *rb_map;
> -   u64 cost;
> -   int err;
>
> if (attr->map_flags & ~RINGBUF_CREATE_FLAG_MASK)
> return ERR_PTR(-EINVAL);
> @@ -174,26 +171,13 @@ static struct bpf_map *ringbuf_map_alloc(union bpf_attr 
> *attr)
>
> bpf_map_init_from_attr(_map->map, attr);
>
> -   cost = sizeof(struct bpf_ringbuf_map) +
> -  sizeof(struct bpf_ringbuf) +
> -  attr->max_entries;
> -   err = bpf_map_charge_init(_map->map.memory, cost);
> -   if (err)
> -   goto err_free_map;
> -
> rb_map->rb = bpf_ringbuf_alloc(attr->max_entries, 
> rb_map->map.numa_node);
> -   if (IS_ERR(rb_map->rb)) {
> -   err = PTR_ERR(rb_map->rb);
> -   goto err_uncharge;
> +   if (!rb_map->rb) {
> +   kfree(rb_map);
> +   return ERR_PTR(-ENOMEM);
> }
>
> return _map->map;
> -
> -err_uncharge:
> -   bpf_map_charge_finish(_map->map.memory);
> -err_free_map:
> -   kfree(rb_map);
> -   return ERR_PTR(err);
>  }
>
>  static void bpf_ringbuf_free(struct bpf_ringbuf *rb)
> --
> 2.26.2
>


[PATCH bpf-next v2 22/35] bpf: eliminate rlimit-based memory accounting for bpf ringbuffer

2020-07-27 Thread Roman Gushchin
Do not use rlimit-based memory accounting for bpf ringbuffer.
It has been replaced with the memcg-based memory accounting.

bpf_ringbuf_alloc() can't return anything except ERR_PTR(-ENOMEM)
and a valid pointer, so to simplify the code make it return NULL
in the first case. This allows to drop a couple of lines in
ringbuf_map_alloc() and also makes it look similar to other memory
allocating function like kmalloc().

Signed-off-by: Roman Gushchin 
---
 kernel/bpf/ringbuf.c | 24 
 1 file changed, 4 insertions(+), 20 deletions(-)

diff --git a/kernel/bpf/ringbuf.c b/kernel/bpf/ringbuf.c
index e8e2c39cbdc9..e687b798d097 100644
--- a/kernel/bpf/ringbuf.c
+++ b/kernel/bpf/ringbuf.c
@@ -48,7 +48,6 @@ struct bpf_ringbuf {
 
 struct bpf_ringbuf_map {
struct bpf_map map;
-   struct bpf_map_memory memory;
struct bpf_ringbuf *rb;
 };
 
@@ -135,7 +134,7 @@ static struct bpf_ringbuf *bpf_ringbuf_alloc(size_t 
data_sz, int numa_node)
 
rb = bpf_ringbuf_area_alloc(data_sz, numa_node);
if (!rb)
-   return ERR_PTR(-ENOMEM);
+   return NULL;
 
spin_lock_init(>spinlock);
init_waitqueue_head(>waitq);
@@ -151,8 +150,6 @@ static struct bpf_ringbuf *bpf_ringbuf_alloc(size_t 
data_sz, int numa_node)
 static struct bpf_map *ringbuf_map_alloc(union bpf_attr *attr)
 {
struct bpf_ringbuf_map *rb_map;
-   u64 cost;
-   int err;
 
if (attr->map_flags & ~RINGBUF_CREATE_FLAG_MASK)
return ERR_PTR(-EINVAL);
@@ -174,26 +171,13 @@ static struct bpf_map *ringbuf_map_alloc(union bpf_attr 
*attr)
 
bpf_map_init_from_attr(_map->map, attr);
 
-   cost = sizeof(struct bpf_ringbuf_map) +
-  sizeof(struct bpf_ringbuf) +
-  attr->max_entries;
-   err = bpf_map_charge_init(_map->map.memory, cost);
-   if (err)
-   goto err_free_map;
-
rb_map->rb = bpf_ringbuf_alloc(attr->max_entries, 
rb_map->map.numa_node);
-   if (IS_ERR(rb_map->rb)) {
-   err = PTR_ERR(rb_map->rb);
-   goto err_uncharge;
+   if (!rb_map->rb) {
+   kfree(rb_map);
+   return ERR_PTR(-ENOMEM);
}
 
return _map->map;
-
-err_uncharge:
-   bpf_map_charge_finish(_map->map.memory);
-err_free_map:
-   kfree(rb_map);
-   return ERR_PTR(err);
 }
 
 static void bpf_ringbuf_free(struct bpf_ringbuf *rb)
-- 
2.26.2